Reinforcement Learning
Geminus
A reinforcement-learning framework for a Splendor AI model, built around self-play, replay data, policy training, and local evaluation.
ZephyrLabs
Where intelligent systems, quantitative research,
and knowledge converge.
Algorithmic TradingAlgo
Algo
Trade
Global market simulation, advanced backtesting, factor research, and execution training. Built for systematic traders.
Research Notes
Journal
A public Obsidian-style knowledge graph distilled from the vault, connecting research, AI, language, philosophy, snowboard notes, and Journal writing.
Todo
- Train an extremely strong Splendor AI. In progress.
- Build a factor-capture quantitative trading model with backtesting validation and live-trading capability.
- Publish a curated Obsidian knowledge graph and distilled public notes.